1. Field of the Invention
The present invention relates to the gradation conversion of a video signal which is suited for adjusting the brightness of an input video signal in a video camera or video printing device.
2. Description of the Prior Art
With recent advances in hard copy printing technologies, and particularly with full-color hard copy printing technologies, it has become possible to faithfully reproduce original images using such printing technologies as subliminal thermal transfer printing. Color reproduction technologies have advanced to the point where colors can be reproduced with silver halide photographic quality depending upon the printing material and image processing technologies applied. By using HDTV and similar high resolution video signals, even the resolution of these reproductions has approached photographic quality.
However, the dynamic range of such printing devices is typically only several multiples of ten between low and high ends thereof, and is even more than ten times worse than that of an average CRT. As a result, it is only possible to reproduce an image comparable to that of the CRT when the dynamic range of the input signal does not exceed the dynamic range of the printer, and the dynamic range of the printer is used to the maximum limit. Automatic gain control (AGC) functions and black level correction functions have therefore been proposed for such printers (Japanese Patent Laid-open Publication No. 64-51890).
Furthermore, a luminance signal offset value is normally provided for adjusting image brightness in a television receiver to effectively adjust the brightness by shifting the luminance signal level. This functions well in a CRT or other device with a wide dynamic range, but in a printer, in which the upper and lower ends of the dynamic range are normally cut off, shifting the luminance signal level to the darker side is limited by the maximum ink density, and by the density of the print medium surface to the light side. Image data is therefore lost, resulting in deteriorated image quality.
To overcome this, one method commonly used changes the gradation characteristics without changing the dynamic range and controls the brightness histogram of the image to visually adjust the image brightness.
As shown in FIG. 13(A), this is achieved by converting the gradation characteristics using gamma conversion of the R, G, B or luminance signals. To make the image brighter, the darkest and brightest parts of the image are left unchanged as shown in the figure while smoothly curving the intermediate levels to the bright side. As a result, the brightness histogram of the image is shifted to the bright side but the dynamic range is not changed, thereby adjusting the image to be visually brighter.
Another application for this technology is in backlight compensation in a video camera.
When recording a backlit subject against a bright background with a video camera, the lens iris is usually opened to admit more light. As shown in FIG. 13(B), the light areas of the background become saturated and washed out (white), eliminating color gradations. As a result, gamma conversion is an effective method of changing color gradations to brighten intermediate tones in backlit photography and video recording.
A conventional gray scale correction method as thus described functions without problems with monochrome images, but results in color hue and saturation also changing when applied to adjust the brightness of color images.
FIG. 16 shows a typical prior art gradation conversion circuit employing gamma conversion circuits for each of the RGB signals, so that R', G' and B' signals are obtained by the following equations: EQU R'=R.sup.g EQU G'=R.sup.g EQU B'=B.sup.g
in which g is variable but is common to all of RGB signals.
FIG. 17 shows another typical prior art gradation conversion circuit employing gamma conversion circuit for luminance signal Y, so that Y' signal is obtained by the following equation: EQU Y'=Y.sup.g
wherein g is variable.
The graphs in FIGS. 14(A) and 14(B) illustrate what happens when conventional gamma conversion gray scale correction is applied to RGB signals, using the circuit shown in FIG. 16, wherein the RGB values are R=0.3, G=0.4, and B=0.5. The case in which the image is made brighter, i.e., the case in which g&gt;1, is shown in FIG. 14(A). Each of the RGB signals is converted, the output values are greater than the input values, and the image is brighter. However, since the value g is the same for the three gamma conversion circuits, such as g=1.2, the rate of increase of RGB signals is different. Thus, the ratio of the output signals R':G':B' differs from the R:G:B ratio of the input signals, resulting in such disadvantages that both hue and saturation are also changed and it is therefore not possible to accurately reproduce the colors of the source image. When the value g becomes great, the drop in saturation is particularly acute because the ratio is closer to 1:1:1.
The case in which the image is made darker, i.e., the case in which g&lt;1, is shown in FIG. 14(B). Each of the RGB signals is converted, the output values are less than the input values, and the image is darker. Again, however, because the input R:G:B and output R':G':B' ratios differ, both hue and saturation are also changed, and because the ratio is increased, there is a noticeable divergence resulting in unnatural saturation.
Thus, when conventional brightness adjustment methods in which the gray scale is changed and the image histogram is changed are applied to color images, both hue and saturation also change.
The graph in FIG. 15 shows the case using the circuit shown in FIG. 17, in which the gray scale correction method is applied to the luminance signal to brighten the image obtained from a video signal consisting of luminance and color difference input signals. As in the previous example, the RGB values are R=0.3, G=0.4, and B=0.5, i.e., Y=0.381 because the luminance signal in the NTSC-format signal is Y=0.3*R+0.59*G+0.11*B; the color difference signals are thus R-Y=-0.081 and B-Y=0.119.
Because this method functions by varying the gray scale characteristics of the luminance signal, it is self-evident that the brightness of intermediate tones can also be changed, but the hue and saturation also change as when the gray scale characteristics of the R, G, and B signals are changed. When the image is made brighter, the value of the converted luminance signal is greater than the input signal and the image is brighter. However because the amplitude of the color difference signal remains unchanged, when the signal is returned to RGB values, R=0.404, G=0.504, and B=0.604; the R:G:B ratio thus differs from that of the original image, and because the ratio approaches 1:1:1, hue changes and saturation drops.
When the image is darkened, the hue likewise changes and the saturation expands unnaturally, but depending upon the input colors negative values may occur when the signal is returned to RGB values.
It has therefore not been possible to adjust just the brightness without changing the color by applying gray scale conversion methods to the luminance signal or separate RGB gray scale correction.